1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
|
/**
* @file llsimdtypes.inl
* @brief Inlined definitions of basic SIMD math related types
*
* $LicenseInfo:firstyear=2010&license=viewerlgpl$
* Second Life Viewer Source Code
* Copyright (C) 2010, Linden Research, Inc.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation;
* version 2.1 of the License only.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*
* Linden Research, Inc., 945 Battery Street, San Francisco, CA 94111 USA
* $/LicenseInfo$
*/
//////////////////
// LLSimdScalar
//////////////////
inline LLSimdScalar operator+(const LLSimdScalar& a, const LLSimdScalar& b)
{
LLSimdScalar t(a);
t += b;
return t;
}
inline LLSimdScalar operator-(const LLSimdScalar& a, const LLSimdScalar& b)
{
LLSimdScalar t(a);
t -= b;
return t;
}
inline LLSimdScalar operator*(const LLSimdScalar& a, const LLSimdScalar& b)
{
LLSimdScalar t(a);
t *= b;
return t;
}
inline LLSimdScalar operator/(const LLSimdScalar& a, const LLSimdScalar& b)
{
LLSimdScalar t(a);
t /= b;
return t;
}
inline LLSimdScalar operator-(const LLSimdScalar& a)
{
static LL_ALIGN_16(const U32 signMask[4]) = {0x80000000, 0x80000000, 0x80000000, 0x80000000 };
ll_assert_aligned(signMask,16);
return _mm_xor_ps(*reinterpret_cast<const LLQuad*>(signMask), a);
}
inline LLBool32 operator==(const LLSimdScalar& a, const LLSimdScalar& b)
{
return _mm_comieq_ss(a, b);
}
inline LLBool32 operator!=(const LLSimdScalar& a, const LLSimdScalar& b)
{
return _mm_comineq_ss(a, b);
}
inline LLBool32 operator<(const LLSimdScalar& a, const LLSimdScalar& b)
{
return _mm_comilt_ss(a, b);
}
inline LLBool32 operator<=(const LLSimdScalar& a, const LLSimdScalar& b)
{
return _mm_comile_ss(a, b);
}
inline LLBool32 operator>(const LLSimdScalar& a, const LLSimdScalar& b)
{
return _mm_comigt_ss(a, b);
}
inline LLBool32 operator>=(const LLSimdScalar& a, const LLSimdScalar& b)
{
return _mm_comige_ss(a, b);
}
inline LLBool32 LLSimdScalar::isApproximatelyEqual(const LLSimdScalar& rhs, F32 tolerance /* = F_APPROXIMATELY_ZERO */) const
{
const LLSimdScalar tol( tolerance );
const LLSimdScalar diff = _mm_sub_ss( mQ, rhs.mQ );
const LLSimdScalar absDiff = diff.getAbs();
return absDiff <= tol;
}
inline void LLSimdScalar::setMax( const LLSimdScalar& a, const LLSimdScalar& b )
{
mQ = _mm_max_ss( a, b );
}
inline void LLSimdScalar::setMin( const LLSimdScalar& a, const LLSimdScalar& b )
{
mQ = _mm_min_ss( a, b );
}
inline LLSimdScalar& LLSimdScalar::operator=(F32 rhs)
{
mQ = _mm_set_ss(rhs);
return *this;
}
inline LLSimdScalar& LLSimdScalar::operator+=(const LLSimdScalar& rhs)
{
mQ = _mm_add_ss( mQ, rhs );
return *this;
}
inline LLSimdScalar& LLSimdScalar::operator-=(const LLSimdScalar& rhs)
{
mQ = _mm_sub_ss( mQ, rhs );
return *this;
}
inline LLSimdScalar& LLSimdScalar::operator*=(const LLSimdScalar& rhs)
{
mQ = _mm_mul_ss( mQ, rhs );
return *this;
}
inline LLSimdScalar& LLSimdScalar::operator/=(const LLSimdScalar& rhs)
{
mQ = _mm_div_ss( mQ, rhs );
return *this;
}
inline LLSimdScalar LLSimdScalar::getAbs() const
{
static const LL_ALIGN_16(U32 F_ABS_MASK_4A[4]) = { 0x7FFFFFFF, 0x7FFFFFFF, 0x7FFFFFFF, 0x7FFFFFFF };
ll_assert_aligned(F_ABS_MASK_4A,16);
return _mm_and_ps( mQ, *reinterpret_cast<const LLQuad*>(F_ABS_MASK_4A));
}
inline F32 LLSimdScalar::getF32() const
{
F32 ret;
_mm_store_ss(&ret, mQ);
return ret;
}
|
